A unified account of null pronouns in Korean

Abstract

This dissertation examines the various manifestations of null pronouns in Korean and attempts to provide a unified account of their properties. Base-generated empty pronouns (pro) and the Generalized Control Rule (GCR: an empty pronominal is coindexed with the closest c-commanding NP), proposed by Huang (1991), playa major role in my explanation. Chapter II examines topic constructions and relative clauses. We will see that in Korean certain phrases cease to be islands in sentence-initial position (as in Chinese), and that Huang's (1991) GCR, with some modification, can be successfully applied to Korean. I revise the GCR by proposing the Matching Condition and the Binding Principles as crucial constraints on the GCR rather than giving up on apparent structural conditions and attributing everything to pragmatic factors. Chapter III examines inalienable possession constructions with 'double' nominative or accusative Case. First, we examine the scope of the 'inalienability' relation that affects the Case alternation. I then propose my analysis, arguing that double nominative/accusative sentences in Korean (inalienable) possession constructions are best explained by positing a pro (licensed by the GCR) in the possessor position of the part-NP. We will see that syntactic restrictions on the part-NP (involving relativization, passivization, and scrambling) do not demonstrate the adjunct status of this element. Rather, they reflect the interaction of the Theta Criterion with the GCR. Other restrictions on the part-NP, such as honorification and reflexivization, are due to the Matching Condition: part-NPs are inanimate while both honorification and reflexivization require a human referent. In Chapter IV, based on the observation that the floated quantifier construction can be related to part-whole construction, I propose that the floated quantifier phenomenon can best be captured by positing a pro in the specifier (partitive genitive) position of the Quantifier NP, which is licensed by the GCR. An extra constraint on the GCR, requiring 'case agreement' between the antecedent NP and the Quantifier NP is proposed for Korean FQCs. It will be shown that the occurrence of classifiers/case markers in the Quantifier NPs should be considered more carefully. Finally, Chapter V summarizes the major points and presents the conclusion.h